am: 30a58db2

Change-Id: I0a0d4cc4a0e8f32d3b3407c370389260f864a3e7

For multiple headset connections, it will be increased to support
multiple services(HFP/AV/CT) for each links.

Test: Connect to 4 or more headsets and play music
Bug: 110972513
Change-Id: I56f69e12fb1a4a614c69fa383e9c53675a189fc9

am: 8432caec

Change-Id: I9f87fd712572e9495b5bceac2f7b96985c963a58

am: 631edaef

Change-Id: Ic87f2bca46c7f3313e7ec9c838ba49fbac0441e5

* When trying to connect SCO to non-active and active device back to
  back, the first connection to non-active device must fail gracefully
* Otherwise, BTA SCO state machine may stuck in a irrecoverable state
* Schedule to send AUDIO_CONNECTING state before starting SCO connection
  to avoid race condition between connecting state and connection
  failure

Bug: 111101075
Test: Manual test, make a call and switch path A->speaker->B on phone UI
Change-Id: I9bfa389e04863944d88f2bca49d27f1247fa072b

Bug: 110818719
Test: connect 5 devices, switch to speaker using dialer, switch to
      Bluetooth device using settings, repeat for all devices

Change-Id: I68d8a892c99e106250245c4f56327555a13543ae

am: 31135bb5

Change-Id: Ica7758b35c6739c8c881cb7c1cd0370bd35d9b40

Check whether the RegisterNotification request from CT is valid, and
safely reject invalid request

Bug: 111260441
Test: manually test with RIVA WAND speaker
Change-Id: Ia2edd94e50fa7cb5b61be54f297adf971060a924

am: f9e306a0

Change-Id: I650b690029c133db6dcde8a29238e7cc360cdcbf

am: 80b719d5

Change-Id: I63fa3d67eb96f3e50396bff5d844a9e4b4336cc3

am: 078f3bb8

Change-Id: I293dc741e31950adeebf2f65c49f7449f74db17e

This fixes issue introduced in commit ddf6320b
Registration for Service Changed notifications was
disabled, because there was no good way back then to have
"opportunistic" app. Since opportunistic apps were introduced, and the
code in GATT_ConfigServiceChangeCCC is already marking connection as
opportunistic, reenable notification registration.

Test: manual verification, that Service Changed descriptor is set
Change-Id: Iefaec3914829fa4a4a60dc3c3790c3daa9f4f4d8

We are sending LE fixed channel notification to classic callback. Stop
doing that. This triggers following warning in log output on every
LE connection:
W bt_smp  : smp_br_connect_callback is called on unexpected transport 2

Test: manual with phone, verify warning is gone when connecting over LE
Change-Id: I47597fa03744fb1bff03d287f790ea32fa033cc7

Add a missing ack in A2DP start failed case when A2DP offload is
enabled.

Bug: 111207927
Test: Manual

Change-Id: I96a7ad322ea7ce6da5084a48fdd3dec980ce57b8

am: 28faf875

Change-Id: Ibdb89a81251633762771e55c00b2f1f492b14094

If the remote device sent us AVDTP Suspend, and then AVDTP Start, don't
trigger AVDTP Suspend from the local device in response to the AVDTP Start.

Also:
 * Add a call btif_av_stream_start_offload() inside btif_a2dp_on_started()
   if we were started remotely.
 * Fix the order of calling btif_a2dp_on_started() and triggering
   an internal BTIF_AV_SUSPEND_STREAM_REQ_EVT event. Otherwise,
   BTA Suspend might be called before btif_av_stream_start_offload()

Bug: 110015851
Bug: 110186160
Test: Manual
Change-Id: I908d19f8caf5130b96a64153ff39bd4b52879713

am: d7e8cd7a

Change-Id: I62a4c95c968ae3476052ae555efa0ad078cf32cf

am: 3f88e806

Change-Id: I71515bfc27707b0659846965148c66b9d535745b

Implement HandleSetAddressedPlayer() to check if remote set valid
player ID and properly respond to remote.

Bug: 110632860
Test: PTS Test AVRCP/TG/MPS/BI-01-C AVRCP/TG/MPS/BV-02-C
Change-Id: Ice13e7ea97cdac369c4d021171a24bf9d96c9da8

am: fdb3552c

Change-Id: Ie853b1c2c21848fa6da57c3db8b0dd59229725a5

Test: Toggle Bluetooth, ls -alZ /vendor/bin/hw/*blue*
Change-Id: Ibf72f8bacc2f4975d06f4892e3b33ef6fb08aab7
Merged-In: Ibf72f8bacc2f4975d06f4892e3b33ef6fb08aab7

am: ddabcc7d

Change-Id: I62b171d4f71ca81e36c06312f93b817482ecf615

dump closes a file descriptor that is still owned by a
ParcelFileDescriptor which leads eventually to a double close.

Test: treehugger
Change-Id: Id351b5429cb17b48e37313850a363d739d78c4a1

If command_timed_out() is triggered and we cannot acquire the
commands_pending_response_mutex to print the debug info, then proceed
with the rest of the operations until the process aborts.

Similarly, if startup_timer_expired() is triggered and we are blocked
by a deadlock, just abort() because there is no way to recover.

Bug: 110441099
Test: Code compilation
Change-Id: I367332bbbd9c1d887bd2bc829b9f617114fc96fc

Merge "resolve merge conflicts of a5a4e276 to stage-aosp-master" into stage-aosp-master

Change-Id: I9ef67219c4a054aa9c2905ba96e6f5e2afb9a493

Test: I solemnly swear I tested this conflict resolution.
Bug: None
Change-Id: Id88890e2e87d94d4bd2a55ac3812d1b11a78b26b

Also clean up the code:
* remove spurious logging, or replace with DVLOG
* use std::copy and std::reverse_copy instead of ARRAY_TO_STREAM and
  REVERSE_ARRAY_TO_STREAM

Test: CryptoToolboxTest already cover all those functions
Change-Id: I3221d0966d2dd6ca229adf441ceb75dd37a54d7c
Merged-In: I3221d0966d2dd6ca229adf441ceb75dd37a54d7c

am: f72bb8b1

Change-Id: I12094731788806e14e0cc57af3b239415e69770e

am: 88c2ae57

Change-Id: I0531174e25223961caaeaa62b45e1ef53142eace

am: 406e1330

Change-Id: I91be0419a5de60d40ca9cc22fedf74573cb2db59